SELECTION FOR IMPROVING SOME YIELD TRAITS IN COTTON

Abstract

The present study was carried out at Sakha Agricultural Research Station, Cotton Research Institute, Agricultural Research Center, Egypt, during 2017 to 2019 seasons. Seven selection procedures (Application 1 to 7) i.e. direct selection for three separately traits (lint cotton yield/plant, bolls/plant and boll weight), Multiplicative index of Subandi et al (1973) involved lint cotton yield/plant, bolls/plant and boll weight and three selection index involved lint cotton yield/plant, bolls/plant and boll weight, lint cotton yield/plant, bolls/plant, boll weight, seed index, lint percentage, lint index and all the studied traits (yield and fiber traits), respectively, to improve lint yield, yield components and fiber properties in early segregating generations; F2, F3 and F4 of a cotton cross (Egyptian variety Giza 87 x Australian genotype 10229). Most of yield traits means in F4 generation were higher than F3 generation via the genetic improvement using the selection procedures. Fiber traits means in F3 and F4 generations were almost the same in values which reflect the early fixation of most of the genetic components of these traits. PCV and GCV for lint cotton yield/plant, bolls/plant and boll weight were larger in F2 generation than those of the succeeding generations. Most of the studied traits showed moderate to high heritability in broad sense in all generations except for bolls/plant and boll weight in F4 generation. Most of fiber traits showed higher heritability in F3 and F4 generations than F2 one. Genotypic correlations between lint cotton yield/plant and bolls/plant in the three generations were positive and highly significant. In F4 generation boll weight showed significant and positive genotypic correlation with almost all the studied traits. Besides, fiber length showed the same trend with boll weight, seed index, lint index and lint percentage. In the direct selection for lint cotton yield/plant and bolls/plant the highest predicted and realized gains from all generations were obtained with direct selection for lint cotton yield/plant and direct selection for bolls/plant in both applications which ranged from 17.43 % with bolls/plant in F4 to 90.49% with lint cotton yield/plant in F2 to application 1 and 2. Direct selection for boll weight (application 3) could increase itself and seed index, lint index, micronaire reading and fiber length. The predicted and realized gains in application 5 and 6 were positive and relatively high for lint cotton yield/plant and bolls/plant in the three generations; and were positive and slightly high for lint percentage, lint index, micronaire reading and fiber length in F4 generation. The predicted and realized gains in application 6 for lint cotton yield/plant and bolls/plant were relatively high in the three generations this mean selection for all yield traits could improve lint cotton yield/plant and bolls/plant by 32.7% and 24.61% in F4 generation. Other yield traits were slightly improved applying this selection procedure, this improvement ranged from 1.05 % for seed index to 8.44% for lint index. Fiber traits were also slightly improved except micronaire reading and uniformity index. Application 7 showed improvements for all the studies traits except micronaire reading. The improvement for lint cotton yield/plant and bolls/plant were not relatively high as the same in the other selection procedures, however boll weight and lint index were relatively high, and the lowest ones were for fiber traits. Four families released from these seven selection applications in F4 generation combined lint yield and most of favorable fiber traits and exceeded the F4 generation mean. These families could be continued to further generations as breeding genotypes for developing higher yield and fiber